Sanity Check on unit consistency for Steady State Thermal Analysis

[casipancmonster]

I have a good many years of working with Abaqus but mostly with structural, quasi-static analysis etc but not much thermal. Right now I have a more or less simple thermal model that I am trying to correlate to measured performance and I fear my unit consistency may not be correct for some of the more unusual properties.

I work in metric: (mm, N, tonne, sec, MPa) which gives mW, mJ etc for thermal.

Does the following look correct?
Length mm
Force N
Mass tonne (10^3kg)
Time s
Stress MPa
Energy mJ
Density tonne/mm^3
acceleration mm/s^2
Power mW
conductivity mW/(mm*K)
specific heat mJ/tonne*K
convection mW/(mm^2*K)
film coefficent mW/(mm^2*K)

for my inputs for something like stainless steel I get:
conductivity of 20
density of 8E-009
elastic 200000
Expansion 2.2E-005
Specific Heat of 4.5E8
film factor (natural convection heat transfer) around 0.0015

Thanks

Luke

 

[IRstuff]

Who came up with this collection of units anyway? I've never gotten any meaningful image from tonne/mm^3. A "nano-tonne" is just weird. Seems like someone went out of their way to pick a really huge mass unit and coupled it with a really small dimension unit, albeit, they could have gone with microns or nanometers, I suppose.

Your convection coefficient seems low. 7.5 W/m^2-K is probably more realistic in air, and partially accounts for radiative loss as well.

TTFN
I can do absolutely anything. I'm an expert!
faq731-376 forum1529

 

[casipancmonster]

The unit system is standard here anyway and make perfect sense for structural analysis. Your forces are in Newtons and stresses are in MPa with mm dimensions. Far more logical than any other way to skin that cat. The density is a bit strange but that is just a material property and not something that is interesting for results. I can't remember the system I used at Boeing in the States but I am sure not being metric it was far more strange. It might be there there is a better way to go for thermal analysis but it is just a matter of getting the order of magnitude correct once and then it is good to go. The issue here is that for thermal with mW per tonne is a bit crazy looking

L